[Coco] MC-10 Questions...

[KnudsenMJ at aol.com]

In a message dated 8/8/04 1:40:23 PM Eastern Daylight Time, jdaggett at gate.net 
writes:

> Since the joystick port measures voltage across a resister,you need to know 
>  the resistance or the current through the resister. 

Actually, the joystick port measures voltage.  Period.  The joysticks 
themselves contain pots (variable resistor taps) that convert a fixed +5V into a 
lesser voltage.

>  To setup the joystick port to read a resistance value you need to measure 
> the 
>  voltage drop across an unknown resister that has as its current source a 
> known  constant current.

True.  What I think the Apple did was to include half of the voltage divider 
internally, and the joystick port (wired as a rheostat, or single-ended 
variable resistor) formed the other half.  The Coco's scheme is more elegant and 
universal.

>  The real problem is that the Coco3 does not really measure the voltage. It 
>  programs a middle value, between 0 and 63,  into a DAC and the joystick 
port 
>  and the output of the DAC are fed into a comparator. To determine the 
exact 
>  position, the DAC values are adjusted until the comparator output changes 
>  state. Kind of an indirect method.

Actually this successive-approximation A/D method is used in just about every 
digital measuring instrument.  Nothing indirect about it.
ISTR that RSBASIC's JOYSTK function does this several times in a row until it 
gets the same reading twice.  Sort of like how I add up my income tax or golf 
score -- as soon as one sum repeats itself, I write that down and quit :-)

> Besides the six bits in the DAC are woefully  very inaccurate.

I think you mean "Imprecise", limited to about 20 mV steps.
Where you get inaccuracy is if the voltage changes while the A/D process is 
homing in on the value.  If the voltage is rapidly changing, or full of 
superimposed noise, the Coco will give you wildly fluctuating readings, just like a 
digital volt-ohm-meter.

That digital dashboard app may need to interpose sample-and-hold circuits to 
hold the voltages steady for the conversion. 
--Mike K.